The impact of herbicides on weed abundance and biodiversity PN0940

147 Pages·2001·1.28 MB·English

The impact of herbicides on weed abundance and biodiversity
PN0940
IACR - Long Ashton Research Station & Marshall Agroecology Ltd
Dr Jon Marshall
Marshall Agroecology Limited, 2 Nut Tree Cottages, Barton Winscombe,
North Somerset, BS25 1DU.
Email: [email protected]
CAER, University of Reading
Professor Valerie Brown
Centre for Agri-Environmental Research (CAER), Department of
Agriculture, University of Reading, Earley Gate, P.O. Box 236, Reading,
RG6 6AT.
Email: [email protected]
CSL, York

Dr Nigel Boatman
Central Science Laboratory, Sand Hutton, York, YO41 1LZ.
Email: [email protected]
IACR – Rothamsted
Dr Peter Lutman
IACR- Rothamsted, Harpenden, Hertfordshire, AL52JQ.
Email: [email protected]
Scottish Crops Research Institute
Dr Geoff Squire
Scottish Crop Research Institute, Invergowrie, Dundee, DD2 5DA.
Email: [email protected]
The impact of herbicides on weed abundance and biodiversity
PN0940
Contents Page
Acknowledgements
EXECUTIVE SUMMARY AND CONCLUSIONS
1. INTRODUCTION 1
1.1. Policy Rationale 1
1.2. Scope of the Desk Study 1
1.3. Objectives 1
1.4. Target and Non-Target Plant (Weed) Species 2
2. ECOLOGY OF REPRESENTATIVE WEED SPECIES 7
3. UPDATING PN0923 - NON-TARGET EFFECTS OF HERBICIDES 13
3.1. Is Biodiversity Important? 13
3.2. Change in Weed Communities 13
3.3. Impacts of Farming 14
3.4. Interactions Between Weed Diversity and Biodiversity 15
3.5. Non-Target Effects Within the Crop 15
3.6. Non-Target Effects Beyond the Crop 15
3.7. Genetically Modified Herbicide Tolerant (GMHT) Crops 16
3.8. Spatial Distribution, Remote Sensing and Mapping of Weeds 16
3.9. Farming Systems 17
4. CHANGES IN ABOVE-GROUND WEED ABUNDANCE 18
5. NON-TARGET WEED SPECIES IN THE SEEDBANK 24
GR Squire, SCRI
5.1. Status Of The Non-Target Species 24
5.2. Abundance And Dynamics Of The Non-Target Species 27
5.3. Suppressive Management – Falling Seedbanks 27
5.4. Relaxing Management – Rising Seedbanks 29
5.5. Community Features And The Potential For Modelling 30
5.6. Dynamics Modelling Linking Trait to Community through 31
Management
5.7. Conclusions 33
6. CURRENT WEED CONTROL AND ITS IMPACT 34
PJW Lutman, IACR
6.1. Introduction 34
6.2. Arable Cropping Patterns 34
6.3. Herbicide Use in 1998 35
6.4. Principal Changes in Herbicide Use Between 1974 And 1998 37
6.5. Changes in Weed Susceptibility to Herbicides Between 1974 and 1998 42
6.6. Changes in Timing of Control of Weeds on Weed Species Diversity 46
PN0940
6.7. Effects Of Fertilisers On Weed Communities 47
6.8. Overall Conclusions 48
7. EFFECTS OF MANAGEMENT OTHER THAN HERBICIDES 62
ON FARMLAND BIODIVERSITY
7.1. Habitat Loss 62
7.2. Fertilisers And Nutrient Enrichment 62
7.3. Cropping Practices 63
8. THE IMPACT OF HERBICIDES ON INVERTEBRATES 64
VK Brown, CAER
8.1. Ecological Framework 65
8.1.1. Invertebrate Communities 65
8.1.2. Attributes Of Weed Communities Important To Invertebrates 65
8.2. Effects of herbicides on invertebrates 66
8.2.1. By Habitat Modification 68
8.2.2. By Prey Resources 70
8.2.3. By Plant Food Resources 71
8.3 New dimensions 79
8.3.1. Effects Of Herbicides On Soil Fauna 79
8.3.2. Sub-Lethal Effects Of Herbicides On Invertebrates 81
9. RELATIONSHIPS BETWEEN WEEDS, HERBICIDES AND BIRDS 90
ND Boatman, CSL
9.1. Introduction 90
9.2. The Diet Of Farmland Birds 91
9.3. Relationships Between Food Abundance And Bird Populations 95
9.4. Relationship Between Breeding Performance And Population Trend 98
9.5. Relationship Between Adult Mortality And Population Trend 99
9.6. Relationship Between Food Density And Foraging 100
9.7. Other Causes Of Bird Declines 101
9.8. Conclusion 102
10. RISK ASSESSMENT FOR NON-TARGET PLANTS 108
WITHIN CROPS
11. PRACTICAL WEED CONTROL, BIODIVERSITY 112
AND RISK AVOIDANCE
11.1. Weed Control And Biodiversity 112
11.2. Risk Management And Avoidance 114
12. KNOWLEDGE GAPS AND RESEARCH NEEDS 117
12.1. Weeds And Weed Management 117
12.2. Invertebrates 120
12.3. Birds 121
12.4. Priority Research Areas 122
EXECUTIVE SUMMARY AND CONCLUSIONS 123
REFERENCES 125
PN0940
Acknowledgements
We wish to thank Dr Lena Ward and the Centre for Ecology and Hydrology for access to
the Phytophagous Insect Database and for useful discussions.
We also wish to thank staff of the Pesticides Safety Directorate for their useful input into
the project, particularly Paul Ashby and Tim Godson. Ingrid Meakin and other staff of the
Chief Scientists Group were also particularly helpful, including finding meeting
accommodation under trying circumstances while moving offices.
We also thank colleagues in our own institutions for their input, without whom this
document would not have been as wide-ranging as it is. This report has been achieved in
just over four months. While there is undoubtedly scope for further work and there are a
number of gaps within the report, it provides a timely review of the impact of herbicides in
the arable ecosystem.
PN0940
EXECTIVE SUMMARY AND CONCLUSIONS
In considering non-target plants within arable fields, the majority of plant species that are
found are of only minor concern to farmers, unless present at high population density.
Under horticultural conditions, it can be argued that all weeds are targets, providing some
difficulty for formal risk assessment. In arable, there are a number of key weed species
that are typically controlled irrespective of density. In contrast, rare arable weeds may
require specific conservation protection; these species may be non-targets under almost all
conditions. The majority of species usually present can be both targets and non-targets and
are most likely to be of greatest significance for biological diversity within fields, as they
occur frequently and with moderate abundance.
Data on farmland birds and invertebrates indicate that there have been significant
reductions in populations and ranges over the past thirty years. In the case of the grey
partridge, there is good evidence that herbicides have played a significant role in their
decline. Whilst habitat loss and fragmentation may play a role in bird declines, the
evidence indicates that habitat degradation is of greater importance. Changes in farming
practice in general are the cause of most population declines of farmland birds. Whilst the
exact causal links are not known for most species, herbicides are implicated.
This review has shown that there have been changes in weed assemblages over the past
century, with some species becoming less common, other increasing in frequency and
others remaining static. Studies of weed seed banks indicate little change in weed seed
abundance or a slight trend for reduced densities. Where weed control has been relaxed,
either as set-aside or where herbicide use has been halved, weed seed banks can increase
rapidly. However, the commonest and most competitive weed species tend to become the
most abundant, under these conditions. Rare species may not recover.
Analysing changes in cropping and herbicide use, the move from spring to winter cropping
since the 1970s has been a dramatic change in cropping practice. Co-incident with the
change to winter cropping, there have been major changes in the pattern of herbicide use.
In the 1970s, herbicides were used primarily for broad-leaved weed control and on only
about 50% of fields. Today, herbicides are used on most fields and are targeted on grass
weeds as well as dicotyledonous species. An examination of the weed spectra controlled
by the herbicides in use over the past 25 years indicates that on average today’s herbicides
control more weeds. Broader spectrum products were introduced in the early 1980s.
Factors other than herbicides may play an important role in changing weed assemblages,
particularly fertilisers and cropping pattern.
Data collected from the literature and from the Phytophagous Insect Database demonstrate
close links between invertebrates and a range of representative weed species. Different
weed species support differing numbers of insect herbivores, with some species hosting
numbers of rare species, as well as pest species. The data indicate that a number of weed
species that are particularly important for insect biodiversity in the arable habitat can be
selected.
Data on the use of weed species by birds has also been examined. Whilst, as with the
invertebrate data, there is some lack of quantitative information on preferences, it is clear
that bird species of conservation importance utilise particular genera of weeds. Thus it is
possible to identify genera that are of greater importance for farmland birds.
PN0940
The data indicate that herbicides, by controlling weeds and modifying abundance and
species assemblages, have impacted on wildlife in arable land. These non-target effects
need to be considered for regulatory reasons, particularly with the requirements under EU
Regulation 91/414. With such dramatic changes in biodiversity, there are also calls for
more sustainable production methods. The challenge will be to grow crops and maintain
an appropriate population of weed species to support farmland wildlife. Under
horticultural conditions, this may be difficult, in terms of crop quality protection.
Nevertheless, under arable and horticultural production, there may be opportunities to
develop sacrifice areas, such as conservation headlands, or to develop much greater
selectivity of herbicide action, either through selective chemistry or application or a
combination of these.
In terms of regulatory needs, the approach of selecting representative weeds and assessing
their importance for biodiversity has been successful. A shortlist of species has been
identified. The approach can now be applied to other weed species, to check the most
important species have been identified. Regulatory approaches reviewed in PN0923 can
be applied as non-target protocols, with adjustment of acceptable risk to achieve control
where required.
There are a number of areas where knowledge is lacking. These are briefly reviewed and a
priority list for research and development is given below:
1. classification of the competitive ability of a wider range of weed species under
different cropping conditions
2. confirmation of the trends shown from data derived from the Phytophagous
Insect Database linking plants to insect herbivores by ecological field study
3. assessment of the biodiversity importance of common weeds not included in
this study
4. surveys of the status of weed and invertebrate populations
5. quantification of the importance of particular weeds for invertebrates and birds,
including preferences and resource values
6. investigation of the interactions between weeds, invertebrate fauna and birds,
including those that are insectivorous at the chick stage
7. modelling the functioning of the agricultural ecosystem to identify clearer
causal links between population change and agronomic practice
8. investigation of the nature and effect of selection pressures within
agroecosystems at genetic, individual, population and community levels
9. development of weed management systems that allow biodiversity to be
maintained in the crop
10. tests of spatial methods of herbicide risk avoidance at appropriate spatial scales
PN0940
1. INTRODUCTION
1.1. Policy Rationale
Herbicides are used to limit reduction in crop yield and quality due to weed competition,
yield contamination and interference with harvesting. Herbicide use has undoubtedly
contributed to crop yield increases and the efficiency of production. However, their
widespread use may have detrimental and unexpected effects on wildlife both within crops
and in associated semi-natural habitats in farmland. DEFRA’s Pesticides Safety
Directorate has a duty to assess risks to non-target organisms as part of its responsibilities
for regulating pesticide use. Aspects of non-target effects of pesticides on terrestrial
wildlife were reviewed in the desk study PN0923, which was completed in 1999.
Developments in assessing risk to non-target organisms since that time have concentrated
on non-target areas, particularly field boundaries, where pesticide drift is likely to occur.
However, significant changes in both population size and population ranges have been
recorded for common bird species of farmland (sustainability indicators) over the past 30
years. There are concerns that significant ecological changes have occurred or are
occurring within arable and horticultural crops associated with herbicide use. Within the
crop, non-crop plants naturally occur. Some of these might be regarded as non-targets.
There is a need to understand the potential direct and indirect effects of herbicides, which
may be mediated by the removal of plant biomass or particular plant species with which
higher trophic taxa are associated, or by affecting processes within soils. What evidence is
there that weed flora have changed with herbicide use? What information is available for
interrogation? How might risks to non-target species be assessed and how might non-
target effects be mitigated?
The study addresses the available information on weed changes, herbicide use patterns and
trophic interactions and non-target effects with invertebrates and birds. In addition, the
conflict between production imperatives and environmental (biodiversity) concerns are
explored, as an attempt to identify approaches to risk assessment within crops and
approaches to practical management of weed flora.
1.2. Scope of the Desk Study
This project examines non-target effects of herbicides on higher plant species within arable
and horticultural crops. It is not concerned with off-target effects, such as drift to semi-
natural habitats, but is particularly concerned with the biodiversity implications of
herbicide use within crops.
1.3. Objectives
The overall objectives of this desk study are to update the review of known effects of
herbicides on weed populations and communities within arable crops and to review the
subsequent indirect effects on fauna, to identify gaps in knowledge, to prioritise research
needs and to examine potential approaches to a) risk assessment for non-target plants in
fields and b) practical means of maintaining appropriate weed cover in crops.
1
PN0940
Specific objectives are to:
1. Define non-target plants in crop situations
2. Review indirect effects of herbicides and other weed management techniques in the
terrestrial environment, building on the review PN0923
3. Examine and evaluate data on the changes in weed communities over the past 50
years
4. Review the relationships between flora and fauna in crop situations
5. Establish nature of current weed control practices and impacts of weeds on arable
crops
6. Define approaches to risk assessment schemes for non-target plants within fields
7. Identify possible and potential approaches to practical weed management that will
satisfy agronomic and wildlife requirements with regard to weed community
structure and abundance
8. Identify gaps in knowledge and prioritise research needs
1.4. Target and Non-target Plant (Weed) Species
Within a crop field, there may be a number of unsown plant species present forming a
weed assemblage. As many of these species compete with the sown crop and reduce yield,
or interfere with harvesting, or contaminate grain samples, farmers and growers regard
them all as weeds worthy of removal, usually by using herbicides. Nevertheless, amongst
these non-crop species, there may be both target and non-target species for weed control.
A number of rare weed species are subject to conservation effort including within
Biodiversity Action Plans (BAPs). These may be regarded as non-target species. Of
greater significance, as they are commoner and often have significant biomass, there is a
suite of species that might be targets at higher density, but non-targets at low population
levels. Finally, there are a number of species that are almost invariably targets for control,
because of their competitive ability and/or their ability to reproduce rapidly.
The weed species that are always targets in arable crops are typically annual grasses, as
well as cleavers (Galium aparine) (Table 1.1). These are particularly associated with
autumn-sown crops, reflecting the predominance of these crops in cultivation.
Table 1.1. Weed species that are almost always targets for weed control if found within
crops.
Species Germination (A = autumn;
S = spring)
Winter wild-oat (Avena fatua) A/S
Spring wild-oat (Avena ssp. ludoviciana) S
Blackgrass (Alopecurus myosuroides) A
Barren brome (Anisantha sterilis) A
Couch grass (Elytrigia repens)
Common cleavers (Galium aparine) A/S
As well as these species, many other species are recorded in arable crops (Jauzein, 1995;
Rodwell, 1995). These may be regarded as both targets and non-targets for weed control,
2
PN0940
depending on a variety of factors. The most important factors affecting the perception of
weeds from the viewpoint of farmers and growers are the relative competitive effect of
particular species and their density. The product of competitive effect and weed density
has been used to estimate crop yield loss (Marshall, 1987; Wilson et al., 1995). Other
factors that will impact on the status of weeds include their effect on harvesting, the purity
of grain samples and their threat to following crops from seed return. Whilst these views
are paramount to farmers, little concern has been paid to the importance of weeds in
general or species in particular for other aspects of the ecological functioning of
agroecosystems. With significant declines reported for a number of plants and animals
associated with farmland, herbicides and weed control may be having rather greater
impacts than hitherto understood. This review addresses both the practical management
requirements for weeds and their role for biological diversity.
At this point, an important caveat to what follows, is required. The review has
concentrated on arable production systems, as these are the largest land uses in the UK.
Nevertheless, horticultural systems are important in certain areas. In these systems,
harvested crop quality is paramount. Therefore, growers would argue that there are no
circumstances under which weed species can be left within the crop as non-targets. There
are “high demands on crop quality and contamination in the horticultural markets, e.g. one
Solanum nigrum (black nightshade) berry found amongst your vining peas and your crop
risks being discarded” (pers comm.. A Grundy, HRI).
In order to evaluate the present state of knowledge in relation to non-target effects of
herbicides and the impact of weed species on agroecosystem function and biodiversity, a
representative list of common weed species has been drawn up. The review has identified
32 common weed species that may or may not be targets for control (Table 1.2). These
species have been selected to represent the spectrum of the following criteria:
1. Frequency: common to less common
2. Competitive ability: economically important to uncompetitive with the crop
3. Environmental value: important to unimportant (so far as known)
4. Taxonomy: representative of main families
The species range from highly competitive to uncompetitive with the crop, with a range of
importance for associated invertebrates and as food for farmland birds. For those plant
species that are often regarded as targets for weed control, many will not need control if
populations are low. For those that are regularly targeted for control, some may be of
particular value for biological diversity. Therefore, the list of plant species given in Table
1.2 includes both target and potentially non-target species. These species are further
examined in terms of their competitive ability, problems for growers, ecology, prevalence,
and their importance for associated animals and birds and ecosystem function.
3
PN0940
Table 1.2. A representative list of common weed species that may be both targets and non-targets for weed control using herbicides.
Common name Latin name Weeds & Weeds & Specialist WMSS Competitive No. for 5% %fields Birds (CSL
birds inverts insects status index yield loss infested review)
Grass weeds
Annual Meadow-grass Poa annua * * priority 0.10 50 79 "
Barren Brome Bromus sterilis * (mice?) * key 13
Black-grass Alopecurus myosuroides * * key 0.40 2-10 38
(mice?+D8)
Wild-oat Avena fatua * key 1.00 42
Broad-leaved weeds
Black Nightshade Solanum nigrum *
Black-bindweed Fallopia convolvulus * * (birds) nd 0.30 "
Broad-leaved Dock Rumex obtusifolius * "
Charlock Sinapis arvensis * (birds) * 0.40 36 "
Cleavers Galium aparine * (mice?) * key 3.00 <1 58
(insects)
Common Chickweed Stellaria media * * (birds, * other 0.20 25 94 "
mice)
Common Field- Veronica persica * * other 0.08 72
speedwell
Common Fumitory Fumaria officinalis * (birds) * 0.08 17 "
Common Hemp-nettle Galeopsis tetrahit * * "
Common Mouse-ear Cerastium fontanum *
Common Poppy Papaver rhoeas * (insects) * priority 0.40 18
Corn Marigold Chrysanthemum segetum nd
Corn Spurrey Spergula arvensis *
Cornflower Centaurea cyanus nd
Creeping Thistle Cirsium arvense * (birds) * 0.30 "
(insects)
Cut-leaved Crane’s-bill Geranium dissectum * 0.08 11
Fat-hen Chenopodium album * * (birds) * priority 0.20 25 13 "
(insects)
Field Forget-me-not Myosotis arvensis * * other 0.20 25
Field Pansy Viola arvensis * * other 0.02 250 45
Fool’s Parsley Aethusa cynapium *
Groundsel Senecio vulgaris * * 0.06 "
4